Machine Learning Methods in the
Environmental Sciences
Neural Network and Kernels
Exercise 1.4
From the given data file (Ex1_4_corr) in Matlab format or text
format, compare the Pearson correlation with the Spearman rank
correlation for the time series x and y (each with 40
observations). Repeat the comparison for the time series
x2 and y2 (from the same data file
as above), where x2 and y2 are the
same as x and y, except that the 5th data point in
y is replaced by an outlier in y2. Repeat the
comparison for the time series x3 and
y3, where x3 and
y3 are the same as x and y, except that
the 5th data point in x and y is replaced by an outlier in
x3 and y3. Make scatterplots of the
data points in the x-y space, the
x2-y2 space and the
x3-y3 space. Also plot the linear
regression line in the scatterplots.
Exercise 1.5
Using the given data file (Ex1_5_MLR) in Matlab
format or text format, perform multiple
linear regression with predictors x1,
x2 and x3, and the response variable
y. Rank the importance of the predictors in their influence on
y.
William W. Hsieh
2009